Les orchid´ees mycoh´et´erotrophes et mixotrophes du Japon

Japon

Une autre grande question concerne l’existence d’esp`eces partiellement myco- h´et´erotrophes (mixotrophes ; Selosse et Roy, 2009) qui sont souvent phylog´en´eti- quement tr`es proches d’esp`eces mycoph´et´erotrophes. `A La R´eunion, on ne connaˆıt pas de proches parents de G. similis qui seraient de bons candidats. Gardant cette recherche pour plus tard sur les Mascareignes, nous avons mis `a profit une colla- boration sur des esp`eces subtropicales avec T. Yukawa (Tsukuba Botanical Garden, Japon) pour commencer `a ´elargir notre connaissance de la strat´egie des orchid´ees mixotrophes hors de la zone temp´er´ee. Au Japon, nous avons ´etudi´e deux esp`eces non chlorophylliennes, Cymbidium macrorhizon et C. aberrans, ainsi que deux esp`eces chlorophylliennes, C. lancifolium et C. goeringii, qui leur sont directement appa- rent´ees dans la phylog´enie du genre Cymbidium (Yukawa et al., 2002).

Cymbidium macrorhizon Cymbidium aberrans Cymbidium lancifolium Cymbidium goeringii Autres Cymbidium spp. Perte de la feuille

Figure2.8 – Positions phylog´en´etiques et photographies des orchid´ees chlorophylliennes et non chlorophylliennes Cymbidium spp. ´etudi´ees au Japon. D’apr`es Yukawa et al. (2002).

Deuxi`eme partie

Publications scientifiques

Chapitre 3

Article 1. Independent

recruitment of saprotrophic fungi

as mycorrhizal partners by

tropical achlorophyllous orchids

R´esum´e Les orchid´ees mycoh´et´erotrophes se sont adapt´ees `a vivre `a l’ombre des forˆets en perdant leur capacit´e de photosynth`ese et en d´ependant de champignons mycorhiziens comme source de carbone. Dans les forˆets temp´er´ees, ces esp`eces sont associ´ees chacune tr`es sp´ecifiquement `a un clade de champignons ectomycorhiziens, exploitant ainsi le carbone organique issu de la photosynth`ese des arbres de la ca- nop´ee avoisinante. Cependant, les communaut´es ectomycorhiziennes sont absentes de la plupart des forˆets tropicales, l`a o`u de nombreuses orchid´ees mycoh´et´erotrophes sont pourtant pr´esentes. Ce constat soul`eve alors la question des strat´egies adopt´ees par les mod`eles tropicaux dans l’´evolution de la mycoh´et´erotrophie.

Nous avons examin´e cette question chez deux esp`eces mycoh´et´erotrophes, qui ne sont pas directement apparent´ees et qui vivent dans des r´egions ´eloign´ees : Gastrodia

similis `a La R´eunion et Wullschlaegelia aphylla en Guadeloupe. J’ai ´echantillonn´e

les populations de G. similis `a La R´eunion avec T. Pailler et J. Fournel de l’Uni- versit´e de La R´eunion. Les populations de W. aphylla ont ´et´e ´echantillonn´ees par M.-A. Selosse et M. Dulormne de l’Universit´e des Antilles et de la Guyane. J’ai r´ealis´e les identifications mol´eculaires des champignons mycorhiziens, ainsi que les broyages des ´echantillons pour les analyses isotopiques, `a Montpellier. J’ai mis des ´echantillons de racines colonis´ees `a la disposition de P. Bonfante et de A. Faccio du CNR de Turin (Italie), qui disposent d’un ´equipement de microscopie ´electronique. Enfin, j’ai analys´e les r´esultats et r´edig´e cet article en interaction avec M-A Selosse

et T. Pailler.

Les deux esp`eces ´etudi´ees ont r´ev´el´e des associations mycorhiziennes `a diff´erents groupes de champignons saprophytes (Basidiomyc`etes). L’esp`ece W. aphylla s’as- socie non sp´ecifiquement `a des myc`enes (Mycena spp.) et `a des marasmes (Gym-

nopus spp.), sans sp´ecialisation locale, alors que l’esp`ece G. similis s’associe plus

´etroitement `a un champignon des Hymenochaetales (Resinicium sp.). L’analyse des signatures isotopiques du carbone et de l’azote chez ces plantes corrobore nos hy- poth`eses sur les sources de carbone : G. similis obtient du carbone indirectement de la d´ecomposition du bois mort, tandis que W. aphylla en re¸coit plus probablement de la d´ecomposition des feuilles mortes au sol. Cette ´etude a r´ev´el´e une diversit´e nou- velle de champignons mycorhiziens, et surtout la strat´egie adopt´ee par les mod`eles tropicaux dans l’´evolution de la mycoh´et´erotrophie. Plus g´en´eralement, elle d´emontre que les organismes peuvent interagir diff´eremment entre les ´ecosyst`emes temp´er´es et tropicaux, et appelle aujourd’hui `a revisiter les interactions biotiques n´eglig´ees dans les r´egions tropicales.

www.newphytologist.org

November 2009

Vol. 184

No. 3

ISSN 0028-646X

• Links between tree

growth and galactic

cosmic rays

• Metal toxicity symptoms

• Genome duplication and

water relations

• Tansley reviews

The CBL–CIPK Ca

2+

_-decoding

signaling network

Microbiological control of

soil-borne diseases

Blackwell Publishing Ltd Oxford, UK NPH New Phytologist 0028-646X 1469-8137

© The Authors (2009). Journal compilation © New Phytologist (2009) 2987 10.1111/j.1469-8137.2009.02987.x July 2009 001??? 14??? Original Article XX XX

Independent recruitment of saprotrophic fungi as

mycorrhizal partners by tropical achlorophyllous orchids

Florent Martos1,2_{, Maguy Dulormne}3,4_{, Thierry Pailler}2_{, Paola Bonfante}5_{, Antonella Faccio}5_{, Jacques Fournel}2_,

Marie-Pierre Dubois1 _{and Marc-André Selosse}1

1_{Centre d’Ecologie Fonctionnelle et Evolutive (CNRS, UMR 5175), Equipe Interactions Biotiques, 1919 Route de Mende, F–34293 Montpellier cedex 5,}

France; 2_{UMR C53 Peuplements Végétaux et Bioagresseurs en Milieu Tropical, Université de La Réunion, 15 Avenue René Cassin, BP 7151, F–97715}

Saint-Denis cedex 9, France; 3_{Conservatoire Botanique des Antilles Françaises, F–97100 Basse Terre, Guadeloupe, France;} 4_{EA 926 DYNECAR, Université des}

Antilles et de la Guyane, BP 592, F–97159 Pointe-à-Pitre, Guadeloupe, France; 5_{Dipartimento di Biologia Vegetale dell’Università, Istituto per la Protezione}

delle Piante – CNR, Viale Mattioli 25, I–10125 Torino, Italy

Summary

• Mycoheterotrophic orchids have adapted to shaded forest understory by shifting to achlorophylly and receiving carbon from their mycorrhizal fungi. In temperate forests, they associate in a highly specific way with fungi forming ectomycorrhizas on nearby trees, and exploiting tree photosynthates. However, many rainforests lack ectomycorrhizal fungi, and there is evidence that some tropical Asiatic species asso- ciate with saprotrophic fungi.

• To investigate this in different geographic and phylogenetic contexts, we identi- fied the mycorrhizal fungi supporting two tropical mycoheterotrophic orchids from Mascarene (Indian Ocean) and Caribbean islands. We tested their possible carbon sources by measuring natural nitrogen (15_{N) and carbon (}13_{C) abundances.}

• Saprotrophic basidiomycetes were found: Gastrodia similis associates with a wood- decaying Resinicium (Hymenochaetales); Wullschlaegelia aphylla associates with both litter-decaying Gymnopus and Mycena species, whose rhizomorphs link orchid roots to leaf litter. The 15_{N and} 13_{C abundances make plausible food chains from}

dead wood to G. similis and from dead leaves to W. aphylla.

• We propose that temperature and moisture in rainforests, but not in most tem- perate forests, may favour sufficient saprotrophic activity to support development of mycoheterotrophs. By enlarging the spectrum of mycorrhizal fungi and the level of specificity in mycoheterotrophic orchids, this study provides new insights on orchid and mycorrhizal biology in the tropics.

Author for correspondence:

Florent Martos Tel: +262 262 938179 Email: florent.martos@univ-reunion.fr Received: 1 April 2009 Accepted: 9 June 2009 New Phytologist (2009) doi: 10.1111/j.1469-8137.2009.02987.x

Key words: mycorrhizas,

mycoheterotrophy, orchids, rainforests, saprotrophic fungi, stable isotopes.

Introduction

Plant–fungal interactions play major roles in terrestrial ecosys- tems by contributing to the structure of plant communities (Smith & Read, 2008) and food chains (Wardle et al., 2004). Although most research on plant–fungal interactions has hitherto been conducted in temperate regions (Alexander & Lee, 2005), there is an increasing interest in tropical ecosystems (Alexander & Selosse, 2009). In addition to playing crucial roles as carbon sinks and in regulating global climate, tropical forests are the Earth’s biologically richest ecosystems (Butler & Laurance, 2008), as species richness is inversely correlated with latitude for many organisms such as plants (Hillebrand, 2004)

and fungi (Öpik et al., 2006; Arnold & Lutzoni, 2007). Beyond the diversity of species, plant–fungal interactions in tropical regions show some differences compared with temperate ones. For example, leaves of tropical plants are hotspots for endophytic fungi, which are more specific (Arnold & Lutzoni, 2007) and have stronger protective effects against plant patho- gens (Arnold et al., 2003) than temperate endophytes. Major differences also occur in mycorrhizal symbiosis, in which plant roots and soil fungi establish a common, dual organ called a mycorrhiza for nutrient exchange (Smith & Read, 2008). In temperate forests, ectomycorrhizal (ECM) symbiosis involving ascomycetes and basidiomycetes predominates among trees. The ECM fungi are rare or missing in many rainforests, where

Research

most trees form arbuscular mycorrhizal (AM) symbiosis with glomeromycetes (Smith & Read, 2008). The AM symbiosis also occurs in temperate forests, but AM fungal communities differ and are more diverse in the tropics (Öpik et al., 2006). The lack of ECM fungi in many rainforests raises questions about tropical mycoheterotrophic (MH) plants. Having evolved independently in several plant lineages, MH species have adapted to shaded forest understory by shifting to achlorophylly (Leake, 1994) and receiving carbon from their mycorrhizal fungi over their whole life cycle. In temperate forests, tens of MH species have been investigated in the last decade (Taylor et al., 2002; Bidartondo, 2005; Roy et al., 2009). They show high mycorrhizal specificity in the sense that each MH species associates with a narrow fungal clade. Moreover, these fungi also form ECM on surrounding trees (Leake, 2004) and give them indirect access to tree photosynthates, as shown for MH orchids (Taylor et al., 2002; Leake, 2004; Roy et al., 2009) and MH Ericaeae (Bidartondo, 2005). Some tropical MH species associate with specific glomeromycete taxa, and thereby connect to nearby AM autotrophic plants (Franke et al., 2006; Merckx & Bidartondo, 2008). The MH orchids do not asso- ciate with glomeromycetes (Rasmussen, 1995; Dearnaley, 2007) and there is evidence, in some tropical Asiatic species at least, that the ecology of associated fungi and the origin of carbon drastically differ.

During in vitro cultivation attempts, saprotrophic or para- sitic fungi were isolated from tropical (incl. subtropical) MH orchids (Rasmussen, 2002), i.e. wood-decaying Erythromyces in Galeola species (Umata, 1995; Dearnaley, 2007), litter- decaying Mycena in Cymbidium (Fan et al., 1996) and patho- genic Armillaria in Gastrodia species (Kusano, 1911; Burgeff, 1932; Kikuchi et al., 2008). However, these works were based on in vitro isolations, where contaminant saprotrophs can overgrow the mycorrhizal partner, or ex-situ mycorrhiza resyn- theses, which can be biased by the absence of suitable partners. Thus, the in situ mycorrhizal status of these fungi was not always confirmed. Recently, molecular approaches revealed Mycena species in Gastrodia confusa (Ogura-Tsujita et al., 2009) and a clade of Coprinaceae in Eulophia zollingeri (Ogura-Tsujita & Yukawa, 2008) and Epipogium roseum (Yamato et al., 2005). Since the latter species can successfully complete its life cycle in in vitro association with isolated Coprinaceae, this sapro- trophic fungus is a plausible partner. Thus, association with ECM or AM fungi does not apply for some MH orchids from tropical Asia, although high specificity seems a general feature of all MH plants.

Some MH species have evolved independently in this family, mainly in rainforests (Leake, 1994), but little is known about tropical MH orchids outside of Asia. In this paper, cellular and molecular approaches, which have not been combined in other studies, demonstrate that saprotrophic fungi form intra- cellular coils (pelotons), making them the mycorrhizal partners of nonAsiatic tropical MH orchids. Furthermore, we support the notion that saprotrophic fungi supply organic matter to MH

orchids by using stable isotope methods. The concentrations of carbon (13_{C) and nitrogen (}15_{N) isotopes enable tracking of}

nutrient sources in ecosystems (Dawson et al., 2002). Organ- isms often have 13_{C abundance similar to that of their food}

source (e.g. MH plants and their associated fungi have very similar composition; Trudell et al., 2003). Forming a conspicu- ous exception, fungi that obtain C from living or decaying plants are less depleted in 13_{C than their substrate (Zeller et al.,}

2007) and, because of this, MH plants are richer in 13_{C than}

surrounding autotrophs. For 15_{N, organisms are usually more}

or less enriched compared with their food source, as demon- strated for fungi (Zeller et al., 2007) and MH plants (Trudell et al., 2003). Thus, 13C and 15_{N contents enable evaluation of}

the plausibility of hypotheses about a given food source. To date, a single study has been performed on tropical MH orchids, and this showed that Gastrodia confusa had similar 13_{C content}

to its saprotrophic Mycena associates, and slightly a higher 15_N

content (Ogura-Tsujita et al., 2009).

Here, we investigate two distantly related tropical MH orchids from the subfamily Epidendroideae, growing in lowland pri- mary rainforests (see the Supporting Information, Table S1) devoid of ECM fungi: the neotropical Wullschlaegelia aphylla (Fig. 1a) from La Guadeloupe, a Caribbean island (Feldmann & Barré, 2001), and the paleotropical Gastrodia similis, an endemic orchid from La Réunion island (Mascarene islands, Indian Ocean; Bosser, 2006; Fig. 1e). The MH abilities arose independently in these two lineages (Molvray et al., 2000), one of which has already been investigated in Asia (Gastrodia: Kusano, 1911; Burgeff, 1932; Wang et al., 1997; Xu & Guo, 2000). We thus investigated two distant regions where no MH orchid had yet been investigated. We show that litter- and wood-decaying fungi, respectively, colonize their roots, and that association is not necessarily highly specific in tropical MH orchids. These observations are contrary to the previous findings in temperate MH plants, and offer new insights into the ecology of plant–fungal interactions in tropical rainforests.

Materials and Methods

Sampling

Wullschlaegelia aphylla (Sw.) Rchb. f. was sampled from four primary rainforests on the island of Guadeloupe (Table 1), during the rainy season, in early December 2007. Two to ten individuals per population were fully harvested and washed. After checking for fungal colonization on a thin section under the microscope, 0.5-mm long infected root fragments were kept for molecular analysis at −80°C (up to 10 per individual). Fungal rhizomorphs sometimes emanated from decaying leaves and coalesced with the roots (Fig. 1c,d): rhizomorphs and root fragments from such points were sampled and kept separately (13 pairs, Table 1). Hyphal pelotons were extracted following Rasmussen (1995) from six root sections (Table 1) neighbouring a 0.5 mm fragment kept for molecular analysis. For each section,

© The Authors (2009) New Phytologist (2009) 184: 668–681

Journal compilation © New Phytologist (2009) www.newphytologist.org

Table 1 Populations of Wullschlaegelia aphylla and Gastrodia similis investigated in this study, with identification of the fungi recovered from the orchid roots (and linked rhizomorphs for W. aphylla)

Location Planta Putative taxonomic identity GenBank accession number Number of samples Putative ecology

W. aphylla Route de la Traversée

(16°10′32″N; 61°41′58″O)

W.a. AV1 Mycenoid sp. 1b _{FJ179470 5 Saprotrophic}

W.a. AV2 Mycenoid sp. 1 FJ179470 3 Saprotrophic

Gymnopoid sp. 1 FJ179475 1 Saprotrophic

W.a. AV2 rhiz. Mycenoid sp. 1 FJ179470 1 Saprotrophic

W.a. AV3 Mycenoid sp. 2 FJ179471 2 Saprotrophic

Gymnopoid sp. 1b _{FJ179475 3 Saprotrophic}

Hypocreales sp. 1 FJ179478 1 Endophyte?

W.a. AV3 rhiz. Mycenoid sp. 2 FJ179471 1 Saprotrophic

W.a. AV3 rhiz. Gymnopoid sp. 1 FJ179475 2 Saprotrophic

W.a. AV4 Gymnopoid sp. 2 FJ179476 1 Saprotrophic

W.a. AV5 Mycenoid sp. 1 FJ179470 3 Saprotrophic

W.a. AV5 rhiz. Mycenoid sp. 1 FJ179470 1 Saprotrophic

W.a. AV6 Mycenoid sp. 1b _{FJ179470 1 Saprotrophic}

W.a. AV6 rhiz. Mycenoid sp. 1 FJ179470 1 Saprotrophic

W.a. AV7 Psathyrella sp. 1c _{FJ179474 2 Saprotrophic}

Hypocreales sp. 1 FJ179478 2 Endophyte?

W.a. AV8 Gymnopoid sp. 2b _{FJ179476 3 Saprotrophic}

Trechisporales sp. 1 FJ179480 1 ?

W.a. AV9 Gymnopoid sp. 1b _{FJ179475 3 Saprotrophic}

Hypocreales sp. 2 FJ179479 1 Endophyte?

Sofaia (16°17′00″N; 61°43′00″O) W.a. FA1 Mycenoid sp. 1 FJ179470 2 Saprotrophic

Gymnopoid sp. 1 FJ179475 2 Saprotrophic

W.a. FA1 rhiz. Mycenoid sp. 1 FJ179470 2 Saprotrophic

W.a. FA2 Mycenoid sp. 3b _{FJ179472 4 Saprotrophic}

Gymnopoid sp. 1b _{FJ179475 1 Saprotrophic}

W.a. FA2 rhiz. Gymnopoid sp. 1 FJ179475 1 Saprotrophic

Tambour (16°09′46″N; 61°38′32″O) W.a. UR1 Mycenoid sp. 3b _{FJ179472 1 Saprotrophic}

Mycenoid sp. 4 FJ179473 1 Saprotrophic

W.a. UR2 Mycenoid sp. 3b _{FJ179472 6 Saprotrophic}

W.a. UR2 rhiz. Mycenoid sp. 3 FJ179472 1 Saprotrophic

W.a. UR3 Mycenoid sp. 2b _{FJ179471 3 Saprotrophic}

Cystofilobasidiales sp. 1 FJ179477 2 ?

W.a. UR3 rhiz. Mycenoid sp. 2 FJ179471 3 Saprotrophic

Contrebandiers

(16°07′01″N; 61°44′22″O)

W.a. CC1 Mycenoid sp. 1b _{FJ179470 2 Saprotrophic}

W.a. CC2 Mycenoid sp. 1 FJ179470 1 Saprotrophic

G. similis Mare Longue

(ML; 21°22′01″S; 55°44′38″E) G.s. ML1 Resinicium sp. 1 FJ179463 3 Saprotrophic Helotiales sp. 1 FJ179465 1 Endophyte? G.s. ML2 Stereaceae sp. 1 FJ179464 3 Saprotrophic G.s. ML3 Resinicium sp. 1 FJ179463 2 Saprotrophic Fusarium sp. 1 FJ179466 1 Endophyte G.s. ML4 Mycenoid sp. 5 FJ179469 2 Saprotrophic Gymnopoid sp. 3 FJ179468 2 Saprotrophic G.s. ML5 Resinicium sp. 1 FJ179463 3 Saprotrophic Trichoderma sp. 1c _{FJ179467 1 Saprotrophic?}

Le Tremblet (21°16′55″S; 55°48′02″E) G.s. TR1 Resinicium sp. 1 FJ179463 2 Saprotrophic

G.s. TR2 Resinicium sp. 1 FJ179463 2 Saprotrophic

Bois Blanc (21°11′44″S; 55°48′33″E) G.s. BO1 Resinicium sp. 1 FJ179463 2 Saprotrophic

G.s. BO2 Resinicium sp. 1 FJ179463 4 Saprotrophic Brûlé Takamaka (21°21′50″S; 55°44′44″E) G.s. KM1 Resinicium sp. 1 FJ179463 1 Saprotrophic G.s. KM2 Resinicium sp. 1 FJ179463 2 Saprotrophic Rivière Sainte-Suzanne (RS; 20°56′57″S; 55°35′00″E) G.s. RS1 Resinicium sp. 1 FJ179463 1 Saprotrophic G.s. RS2 Resinicium sp. 1 FJ179463 1 Saprotrophic G.s. RS3 Resinicium sp. 1 FJ179463 3 Saprotrophic G.s. RS4 Resinicium sp. 1 FJ179463 3 Saprotrophic

a_{‘rhiz.’ indicates analysis of an emanating fungal rhizomorph.}

b_{Pelotons extracted from a root section neighbouring one sample gave the same PCR product.}

c_{Only ITS sequences for Trichoderma sp. 1 and Psathyrella sp. 1 (otherwise, the 5′ part of the 28S rDNA was also sequenced).}

See Supporting Information Table S3 and Fig. 5 for detailed fungal identification. Research

three pools of 15 pelotons each were obtained and kept at −_{80°C for molecular analysis. Sampling for isotopic studies was} conducted at Route de la Traversée in December 2007, and included (n = 7 each): green leaves recently fallen from the forest canopy; orchid tuberous roots (without fungus); brown dead tree leaves; and tree leaves attached to rhizomorphs coalescing with W. aphylla roots. Dead leaves were from uniden- tified tree species. We also sampled fruitbodies of seven litter saprotrophic basidiomycetes (Table S2; n = 2 each) collected in 2007 and inflorescences (n = 7) of W. aphylla collected in May 2006.

Gastrodia similis Bosser was sampled between August and September 2006 in La Réunion from four primary rainforests and one second-growth forest (Rivière Sainte-Suzanne, Table 1). Root fragments were sampled as for W. aphylla (up to five per individual). Sampling for isotopic studies was conducted in Mare Longue (ML) and Rivière Sainte-Suzanne (RS). In each site, we collected (n = 6 each): decaying wood where the orchid was growing, orchid inflorescences, mycorrhizas, dead and living tree leaves of two green tree species (Agarista salicifolia G. Don and Syzygium jambos Alston at RS, and Gaertnera vaginata Poir. and Mimusops balata (Aubl.) C.F. Gaertn.

at ML). In addition, wood-decaying basidiomycetes were collected (three species at ML and one at RS, n = 6 fruitbodies for each; Table S2).

Microscopy

We randomly sampled fragments adjacent to those kept for molecular analysis; some contact points between fungal rhizo- morphs and W. aphylla mycorrhizas were also harvested. Twelve samples for W. aphylla and eight for G. similis were quickly fixed and handled as in Roy et al. (2009) to obtain semithin 0.05 µm sections for transmission electron microscopy and semithin 1 µm sections stained with 1% toluidine blue for light microscopy.

Molecular identification of the fungi

To identify the fungus, we amplified the fungal internal tran- scribed spacer (ITS) plus the 5′ part of the 28S rDNA, using the primers ITS1F and TW13, as described in Roy et al. (2009). To further control the absence of usual orchid sym- bionts, additional PCR amplifications were carried out using

© The Authors (2009) New Phytologist (2009) 184: 668–681

Journal compilation © New Phytologist (2009) www.newphytologist.org

Research 671

Fig. 1 Morphology of Wullschlaegelia

aphylla (a–d) and Gastrodia similis (e–g). (a) Inflorescence of W. aphylla. (b) Ramified mycorrhizal roots (r) and short tuberous roots (t) in leaf litter. (c) Fungal rhizomorphs (rh) linking mycorrhizas and decaying leaves (rhizomorphs are dark on the left and pale on the right). (d) Contact between a rhizomorph and a mycorrhiza. (e) inflorescence of G.

similis. (f) Thick starch-filled rhizomes (rz) and long mycorrhizal roots (r) in dead wood. (g) A rhizome with emanating mycorrhizas.

specific primers ITS4tul for tulasnelloids and ITS3S for sebacinoids, as described in Selosse et al. (2004) and positive controls; no amplification was obtained (not shown). Sequ- encing was performed as described in Roy et al. (2009). To check for contamination, sequences were compared with the fungal sequences obtained in our laboratory since 2004; no similarity was discovered (not shown). Corrected sequences were deposited in GenBank. Sequences were identified by blast _{analysis against GenBank (Table S3) and phylogenetic} placements (Fig. 5).

Isotopic analysis

All samples were dried at 45°C for 48 h and ground in 1.5 ml Eppendorf tubes using 1.1 mm diameter tungsten carbide balls (Biospec Products, Inc., Bartlesville, OK, USA) in a Retch MM301 vortexer (Retch Gmbh and Co., Haan, Germany). Total abundances of 13_{C and} 15_{N were measured using an on-}

line continuous flow CN analyser coupled with an isotope ratio mass spectrometer (Ohlsson & Wallmark, 1999). Isotope abundances are expressed in δ13_{C and δ}15_{N values in parts per}

thousand relative to international standards Vienna-Pee-Dee Belemnite and atmospheric N₂: δ13_{C or δ}15_{N = (R}

sample/

Rstandard– 1) × 1000, where R is the molar ratio (i.e. 13C/12C

or 15_N/14_{N). The standard deviation of the replicated}

standard samples (n = 13) was 0.029‰ for 13_{C and 0.288‰}

for 15_N.

Statistics

Total N, C : N ratio, 13_{C and} 15_{N values were tested for}

normality and homogeneity of variances using a Shapiro–Wilks test and a Levene test, respectively. One-way ANOVAS were performed for each variable and each site, followed by a pairwise t-test (Bonferroni corrected) to calculate pairwise comparisons between group levels at α = 0.01. All values were estimated by mean values followed by 95% confidence intervals (CI 95). Statistical analyses were computed using R 2.7.1 (R Foundation, Vienna, Austria).

Results

Patterns of fungal colonization

Wullschlaegelia aphylla had starch-filled, uncolonized tuberous roots, and long mycorrhizal roots (Figs 1b, S1a) that densely ramified within the leaf litter during the rainy season (Fig. S1b). Mycorrhizas were often connected to decaying leaves by fungal rhizomorphs (elongated bundles of fungal hyphae; Figs 1c,d, S3). The epidermal layer of mycorrhizas was often senescent (Fig. 2a), and some cells contained densely packed hyphae (boxed in Fig. 2b), glued together by an amorphous material (Fig. 2c) and morphologically similar to those forming rhizomorphs. The outer root cortex layer was uncolonized, while polygonal

Dans le document Structuration écologique et évolutive des symbioses mycorhiziennes des orchidées tropicales (Page 61-200)